next up previous contents
Next: 4.3 Capacitance-Voltage Characteristics Up: 4. Characterization of Interfaces Previous: 4.1 Charge Pumping Method


4.2 DCIV Method

Another experimental method for the investigation of the interface quality is the direct-current current-voltage (DCIV) method. It can be used to determine both the interface trap density \ensuremath {N_\textrm {it}} and the amount of fixed oxide charges \ensuremath {Q_\textrm {ox}}.

Figure 4.10: Configuration of the DCIV measurement. The source and drain to substrate diodes are slightly forward biased. As the gate to substrate voltage \ensuremath {V_\textrm {g}} is swept, a current at the substrate contact can be measured.
\includegraphics[width=10cm]{figures/dciv-schematic}
The experimental set-up is illustrated in Figure 4.10. The source and drain to substrate diodes are slightly forward biased, typically with around $\ensuremath{V_\textrm{e}}= \ensuremath{V_\textrm{s}}= \ensuremath{V_\textrm{d}}\approx 0.33 $V. The gate bias is swept from inversion to slight accumulation. During this sweep the substrate current \ensuremath{I_\textrm{b}} is measured. The substrate current originates from recombination of carriers at the \ensuremath {\textrm {Si/SiO$_2$}} interface.

Figure 4.11: DCIV currents after different stress times. With increasing interface degradation the maximum \ensuremath {I_\textrm {DCIV}} increases. The shift of the peak position to lower voltage indicates fixed oxide or interface charges. Data are from Zhu et al. [45].
\includegraphics[width=\figwidth]{figures/zhu05b-dciv}

For each degradation level there is a clear peak of \ensuremath {I_\textrm {DCIV}} at a certain $\ensuremath{V_\textrm{g}}
= \ensuremath{V_\textrm{peak}}$, as shown in Figure 4.11. The peak height above the base line, \ensuremath{\Delta I_\textrm{DCIV}} is approximately proportional to the effective interface trap density \ensuremath {N_\textrm {it}} as [46,47]

\begin{displaymath}
\ensuremath{\Delta I_\textrm{DCIV}}= \frac{1}{2}\ensuremath...
...e}}\vert}{2\ensuremath{\textrm{k$_\textrm{B}$}}T} \right)   .
\end{displaymath} (4.7)

Here, \ensuremath {\textrm{q}_0} is the elementary charge, \ensuremath {n_\textrm{i}} the intrinsic carrier concentration, $\sigma$ the geometric mean of the capture cross sections for electrons and holes, equal to $\sqrt{\sigma_\mathrm{e}\sigma_\mathrm{h}}$, \ensuremath{v_\textrm{th}} the thermal velocity, $A$ the effective gate area, \ensuremath{V_\textrm{e}} the forward bias applied to the source-substrate and drain-substrate junctions, \ensuremath{\textrm{k$_\textrm{B}$}} Boltzmann's constant, and $T$ the temperature.

To determine \ensuremath {Q_\textrm {ox}}, the oxide charge density, the peak position of the DCIV curve can be used. As described in Section 2.2.1, the peak position is directly proportional to the amount of oxide charges and can be calculated using Equation 2.31.


next up previous contents
Next: 4.3 Capacitance-Voltage Characteristics Up: 4. Characterization of Interfaces Previous: 4.1 Charge Pumping Method

R. Entner: Modeling and Simulation of Negative Bias Temperature Instability